ICU 55.1  55.1
unorm2.h
Go to the documentation of this file.
1 /*
2 *******************************************************************************
3 *
4 * Copyright (C) 2009-2013, International Business Machines
5 * Corporation and others. All Rights Reserved.
6 *
7 *******************************************************************************
8 * file name: unorm2.h
9 * encoding: US-ASCII
10 * tab size: 8 (not used)
11 * indentation:4
12 *
13 * created on: 2009dec15
14 * created by: Markus W. Scherer
15 */
16 
17 #ifndef __UNORM2_H__
18 #define __UNORM2_H__
19 
31 #include "unicode/utypes.h"
32 #include "unicode/localpointer.h"
33 #include "unicode/uset.h"
34 
42 typedef enum {
85 
111 
116 struct UNormalizer2;
117 typedef struct UNormalizer2 UNormalizer2;
119 #if !UCONFIG_NO_NORMALIZATION
120 
132 U_STABLE const UNormalizer2 * U_EXPORT2
133 unorm2_getNFCInstance(UErrorCode *pErrorCode);
134 
146 U_STABLE const UNormalizer2 * U_EXPORT2
147 unorm2_getNFDInstance(UErrorCode *pErrorCode);
148 
160 U_STABLE const UNormalizer2 * U_EXPORT2
162 
174 U_STABLE const UNormalizer2 * U_EXPORT2
176 
188 U_STABLE const UNormalizer2 * U_EXPORT2
190 
212 U_STABLE const UNormalizer2 * U_EXPORT2
213 unorm2_getInstance(const char *packageName,
214  const char *name,
215  UNormalization2Mode mode,
216  UErrorCode *pErrorCode);
217 
233 U_STABLE UNormalizer2 * U_EXPORT2
234 unorm2_openFiltered(const UNormalizer2 *norm2, const USet *filterSet, UErrorCode *pErrorCode);
235 
242 U_STABLE void U_EXPORT2
243 unorm2_close(UNormalizer2 *norm2);
244 
245 #if U_SHOW_CPLUSPLUS_API
246 
248 
259 
261 
262 #endif
263 
280 U_STABLE int32_t U_EXPORT2
281 unorm2_normalize(const UNormalizer2 *norm2,
282  const UChar *src, int32_t length,
283  UChar *dest, int32_t capacity,
284  UErrorCode *pErrorCode);
303 U_STABLE int32_t U_EXPORT2
305  UChar *first, int32_t firstLength, int32_t firstCapacity,
306  const UChar *second, int32_t secondLength,
307  UErrorCode *pErrorCode);
326 U_STABLE int32_t U_EXPORT2
327 unorm2_append(const UNormalizer2 *norm2,
328  UChar *first, int32_t firstLength, int32_t firstCapacity,
329  const UChar *second, int32_t secondLength,
330  UErrorCode *pErrorCode);
331 
351 U_STABLE int32_t U_EXPORT2
353  UChar32 c, UChar *decomposition, int32_t capacity,
354  UErrorCode *pErrorCode);
355 
385 U_STABLE int32_t U_EXPORT2
387  UChar32 c, UChar *decomposition, int32_t capacity,
388  UErrorCode *pErrorCode);
389 
405 U_STABLE UChar32 U_EXPORT2
406 unorm2_composePair(const UNormalizer2 *norm2, UChar32 a, UChar32 b);
407 
417 U_STABLE uint8_t U_EXPORT2
419 
436 U_STABLE UBool U_EXPORT2
437 unorm2_isNormalized(const UNormalizer2 *norm2,
438  const UChar *s, int32_t length,
439  UErrorCode *pErrorCode);
440 
459 unorm2_quickCheck(const UNormalizer2 *norm2,
460  const UChar *s, int32_t length,
461  UErrorCode *pErrorCode);
462 
487 U_STABLE int32_t U_EXPORT2
489  const UChar *s, int32_t length,
490  UErrorCode *pErrorCode);
491 
501 U_STABLE UBool U_EXPORT2
503 
513 U_STABLE UBool U_EXPORT2
515 
524 U_STABLE UBool U_EXPORT2
525 unorm2_isInert(const UNormalizer2 *norm2, UChar32 c);
526 
527 #endif /* !UCONFIG_NO_NORMALIZATION */
528 #endif /* __UNORM2_H__ */
const UNormalizer2 * unorm2_getNFDInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFD normalization.
"Fast C or D" form.
Definition: unorm2.h:73
The input string is not in the normalization form.
Definition: unorm2.h:96
const UNormalizer2 * unorm2_getInstance(const char *packageName, const char *name, UNormalization2Mode mode, UErrorCode *pErrorCode)
Returns a UNormalizer2 instance which uses the specified data file (packageName/name similar to ucnv_...
int32_t unorm2_getRawDecomposition(const UNormalizer2 *norm2, UChar32 c, UChar *decomposition, int32_t capacity, UErrorCode *pErrorCode)
Gets the raw decomposition mapping of c.
Map, and reorder canonically.
Definition: unorm2.h:60
UBool unorm2_isInert(const UNormalizer2 *norm2, UChar32 c)
Tests if the character is normalization-inert.
The input string is in the normalization form.
Definition: unorm2.h:101
void unorm2_close(UNormalizer2 *norm2)
Closes a UNormalizer2 instance from unorm2_openFiltered().
UBool unorm2_isNormalized(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Tests if the string is normalized.
const UNormalizer2 * unorm2_getNFCInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFC normalization.
The input string may or may not be in the normalization form.
Definition: unorm2.h:109
Compose only contiguously.
Definition: unorm2.h:83
int32_t unorm2_getDecomposition(const UNormalizer2 *norm2, UChar32 c, UChar *decomposition, int32_t capacity, UErrorCode *pErrorCode)
Gets the decomposition mapping of c.
UNormalizer2 * unorm2_openFiltered(const UNormalizer2 *norm2, const USet *filterSet, UErrorCode *pErrorCode)
Constructs a filtered normalizer wrapping any UNormalizer2 instance and a filter set.
C API: Unicode Set.
UNormalizationCheckResult unorm2_quickCheck(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Tests if the string is normalized.
#define U_NAMESPACE_BEGIN
This is used to begin a declaration of a public ICU C++ API.
Definition: uversion.h:129
int32_t unorm2_normalize(const UNormalizer2 *norm2, const UChar *src, int32_t length, UChar *dest, int32_t capacity, UErrorCode *pErrorCode)
Writes the normalized form of the source string to the destination string (replacing its contents) an...
#define U_DEFINE_LOCAL_OPEN_POINTER(LocalPointerClassName, Type, closeFunction)
"Smart pointer" definition macro, deletes objects via the closeFunction.
Definition: localpointer.h:338
C++ API: "Smart pointers" for use with and in ICU4C C++ code.
struct UNormalizer2 UNormalizer2
C typedef for struct UNormalizer2.
Definition: unorm2.h:117
int32_t UChar32
Define UChar32 as a type for single Unicode code points.
Definition: umachine.h:332
const UNormalizer2 * unorm2_getNFKDInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFKD normalization.
UChar32 unorm2_composePair(const UNormalizer2 *norm2, UChar32 a, UChar32 b)
Performs pairwise composition of a & b and returns the composite if there is one. ...
const UNormalizer2 * unorm2_getNFKCCasefoldInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFKC_Casefold normalization.
uint16_t UChar
Define UChar to be UCHAR_TYPE, if that is #defined (for example, to char16_t), or wchar_t if that is ...
Definition: umachine.h:312
#define U_NAMESPACE_END
This is used to end a declaration of a public ICU C++ API.
Definition: uversion.h:130
struct USet USet
Definition: ucnv.h:67
UNormalization2Mode
Constants for normalization modes.
Definition: unorm2.h:42
UErrorCode
Error code to replace exception handling, so that the code is compatible with all C++ compilers...
Definition: utypes.h:476
uint8_t unorm2_getCombiningClass(const UNormalizer2 *norm2, UChar32 c)
Gets the combining class of c.
int32_t unorm2_normalizeSecondAndAppend(const UNormalizer2 *norm2, UChar *first, int32_t firstLength, int32_t firstCapacity, const UChar *second, int32_t secondLength, UErrorCode *pErrorCode)
Appends the normalized form of the second string to the first string (merging them at the boundary) a...
UBool unorm2_hasBoundaryAfter(const UNormalizer2 *norm2, UChar32 c)
Tests if the character always has a normalization boundary after it, regardless of context...
int32_t unorm2_spanQuickCheckYes(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Returns the end of the normalized substring of the input string.
Basic definitions for ICU, for both C and C++ APIs.
UBool unorm2_hasBoundaryBefore(const UNormalizer2 *norm2, UChar32 c)
Tests if the character always has a normalization boundary before it, regardless of context...
int32_t unorm2_append(const UNormalizer2 *norm2, UChar *first, int32_t firstLength, int32_t firstCapacity, const UChar *second, int32_t secondLength, UErrorCode *pErrorCode)
Appends the second string to the first string (merging them at the boundary) and returns the length o...
Decomposition followed by composition.
Definition: unorm2.h:51
"Smart pointer" class, closes a UNormalizer2 via unorm2_close().
const UNormalizer2 * unorm2_getNFKCInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFKC normalization.
UNormalizationCheckResult
Result values for normalization quick check functions.
Definition: unorm2.h:91
#define U_STABLE
This is used to declare a function as a stable public ICU C API.
Definition: umachine.h:109
int8_t UBool
The ICU boolean type.
Definition: umachine.h:234